KIMSUFI MONITOR LAMBDA

This lambda function is used to monitor the OVH server availability API for a list of servers and send a notification through a slack web hook if one or more servers are available.

Configuration file

{
  "servers": {
    "160sk1": "Kimsufi KS-1 (HDD 500Gb)",
    "160sk2": "Kimsufi KS-2 (HDD 1Tb)",
    "160sk21": "Kimsufi KS-2 (SSD 40Gb)",
    "160sk22": "Kimsufi KS-2 (SSD 80Gb)"
  },
  "slack": {
    "bot_url": "<YOUR SLACK HOOK HERE>",
    "bot_name": "OVH Servers",
    "bot_message_tpl": "{server} is available!"
  }
}

Before packaging your function you need to create a config.json file to store:

• what servers are you interested in
• the webhook url for your slack bot. See the slack documentation for more information about webhooks and slack.

You can copy config-sample.json and use it as a starting point for your own configuration file.

Packaging

It's easy. Just run:

make

This will download all the dependencies and generate a kimsufi-monitor.zip file you can use in your lambda function.

Deploying

There are at least a couple of ways to deploy an AWS lambda function, but for this kind of small upload-once-and-forget-about-it functions I like the CLI more. Obliviously the following commands assume you already installed and configured aws-cli on your current host (see the documentation if it is not) and you have some familiarity with AWS terminology.

Let's then create our new function!

aws lambda create-function \
  --function-name KimsufiMonitor \
  --runtime python2.7 \
  --memory 128 \
  --timeout 10 \
  --handler main.handle_lambda \
  --role "<your lambda execution role ARN>" \
  --zip-file fileb://./kimsufi-monitor.zip \
  --description "Kimsufi Monitor" --output json

This should return to you something like (depending in what version you are downloading from this repository):

{
    "Description": "Kimsufi Monitor",
    "Runtime": "python2.7",
    "CodeSize": 950867,
    "FunctionName": "KimsufiMonitor",
    "CodeSha256": "b/gcaDsaL5ly13pMRs301IG8BMdWGTUBcHfbwC/tsiQ=",
    "Handler": "main.handle_lambda",
    "Timeout": 10,
    "Role": "arn:aws:iam::0000000:role/lambda_kimsufi_monitor",
    "FunctionArn": "arn:aws:lambda:us-east-1:0000000:function:KimsufiMonitor",
    "LastModified": "2016-08-02T21:25:23.417+0000",
    "MemorySize": 128,
    "Version": "$LATEST"
}

This is unfortunately not enough. We have a lambda function yes, but can only execute it manually right now, and instead we want to schedule it. To do that, we need to leverage CloudWatch events, like below.

aws events put-rule \
  --name "1HourRule" \
  --schedule-expression "rate(1 hour)" --output json
aws events put-targets \
  --rule "1HourRule"
  --targets "Id=km,Arn=arn:aws:lambda:us-east-1:0000000:function:KimsufiMonitor"

For the second command, in the targets parameter, the Arn to use is the FunctionArn we got when creating the lambda function, while the Id is simply a unique string.

And we are all set!

Enjoy your notifications, I guess.